That rabbit--fat has 9 calories per gram, fat-free lean meat works out to more like a calorie per gram (you'll see very lean chicken breast. So you could have meat with 10 percent fat and still have that be half the calories.

This bit from the article;



One problem here is that 30 calories is outside of the measurement accuracy available to us, most of the time. If I measure out 2000 calories worth of various foods on my electronic food scale, I doubt I'll be anywhere near that accurate. With calorie restriction--whatever might work, we're forced to make changes large enough to be certain that we actually changed something. Sort of the same problem with weighing ourselves, if it turned out the best way to lose weight was an ounce a week, we'd be out of luck, we could never be certain anything was happening. So--it's actually very hard to know that you've eliminated the equivalent of two or three M&Ms. Of course if you do manage to, calorie requirements go down as you lose weight, so you don't get to goal that way anyways. A 30 calorie a day surplus might get people fatter, but they'd have to increase, account for their increased metabolic rate as they get bigger, or they'll just reach an equilibrium where they're burning off the extra 30 calories. I think more likely--people will be eating a greater surplus. Maybe they go to college and start cooking for themselves, maybe shifting from puberty to young adult hood changes something--more likely than a prolonged slight surplus is something like say an extra 3 or 4 hundred calories a day, and the metabolism eventually catches up to that. Or maybe that scenario just seems more believable to me because at least it's within measurement accuracy, sort of an observability fallacy.